Artificial intelligence and machine learning in autonomous driving: current trends and future directions: a review

Engineering notes

Key topics: autonomous driving, object detection, sensor fusion, reinforcement learning, perception, planning, control. See the paper for implementation details and experimental results.

Chinese explanation / 中文解读

中文解读待补充：本站会优先为端到端自动驾驶、BEV感知、3D目标检测、轨迹预测、路径规划、LiDAR感知等高价值论文补充中文说明。

Original abstract

Abstract Artificial Intelligence (AI) and Machine Learning (ML) have revolutionized autonomous driving (AD), enabling vehicles to perceive, reason, and act with unprecedented precision in dynamic environments. This review synthesises advancements across core AD domains, such as perception systems that leverage convolutional neural networks (CNNs) and multimodal sensor fusion techniques to improve object detection accuracy. Advancements in complex traffic collision-free navigation using reinforcement learning (RL), hierarchical planning, and the adoption of model predictive control (MPC) and adaptive algorithm applications to reduce tracking errors in autonomous driving were considered in this review. However, critical challenges persist that affect the efficiency of AD systems, such as data scarcity, adversarial vulnerabilities, ethical dilemmas in life-critical scenarios, and inconsistent regulatory frameworks. Emerging solutions such as edge computing, physics-aware sensor fusion, and explainable AI (XAI) promise to enhance real-time reliability and public trust. This review determined that future progress hinges on cross-disciplinary collaboration to address technical robustness, societal equity, and global standardization. Therefore, charting a roadmap bridging AI innovation and ethical governance for safety, scalability, and socially responsible autonomous mobility.

Links and sources

• PDF from original source

Comments